The only thing more painful than having a book one champions continually rejected is the practice of self-censorship one must engage in from time to time just as a matter of expediency. Some books that I find fascinating, I know from experience, most editors won't touch.
It pains me to think of all the average, but marketable, books that get published every year, and all the weird, but deliciously good, ones that don't. If you're curious about how this fascinating selection process works, I'd recommend taking a look at my colleague Denise Little's blog post about the way books are acquired and published entitled Two or More Years Before Publication: The Publisher at Work.
To summarize: the acquisitions process is a group process involving the editorial staff, as well as marketing, sales, and the art departments. Books are chosen by popular vote and evaluated not just for their artistic merits, but for their saleability. Lots of funny things can happen in these meetings that can prevent a book from making it to publication. Often times editors are crestfallen that their pet project didn't survive the scrutiny of their co-workers. I've been on the receiving end of many a regretful rejection letter from an editor who, as much as they loved the book, failed to get it picked up at their acquisitions meeting. Separate from the acquisitions process, some imprints also have editorial meetings, wherein they winnow down selections before even presenting them for acquisition.
There are many ways that this process is effective from a practical standpoint. Group decisions help build coalitions and foster a sense of group responsibility that will guide a book through each stage in the process of production and sales. The hope is that these group decisions will more closely mirror the current zeitgeist, and that the books chosen will therefore reach the maximum number of readers. This is sometimes the case, but the downside of these group decisions, in my opinion, is that they create a culture of self-censorship which then trickles down to agents like me. Editors don't want to bring oddball projects to their editorial or acquisitions meetings if they know they're going to get shot down, and so they avoid them altogether.
I don't believe that there's any special talent editors or agents possess that make them any better in their role as taste-makers than your average reader. The decisions they make about which books to keep and which books to toss, are similar to the type of decisions shoppers make at bookstores every day. No one likes to be bored, and editors and agents are no exception. The one thing that differentiates them from other readers is the surrounding cultural influences of the industry itself, and the narratives constructed around the successes or failures of the books they publish. While publisher's like to think they're giving their customers what they want, in truth it is a publishing professional's preconceived notions about what a commercial book should look like that most influences what readers end up with.
The lessons a publisher extrapolates from the data compiled on a book can be rather unscientific. One poor showing can mean writing off an entire sub-genre and some successes are falsely attributed to the popularity of a theme when they're really based on the popularity of an author (or visa versa). In leaner times, like the ones we're experiencing now, this sort of problem only gets exacerbated. On the one hand publishers are being very cautious about trying different sorts of books, and on the other they are looking to throw their weight behind previous successes that are rapidly approaching their expiration date. People tend to respond to scarcity by being more conservative, and that can be a recipe for stagnation. Is it possible that in our efforts to better serve existing readerships that we're limiting the opportunities to connect with new readers, and to open the door to new styles and genres?
What are we so afraid of? There are several admirable books that pass muster with the acquisitions team, but which flop anyway. No one has a complete lock on the marketplace, and bubbles are bursting all around us. What would be the harm in taking a crazy risk once a season, rather than once in a career? If your bubbles are bursting, why try to prevent the others from bursting when you can just blow more bubbles?
I propose that imprints seriously consider publishing one hare-brained book a season, and market it as such. Restaurants have specials every day as a way of experimenting with which dishes to add to the menu, why can't imprints create a slot each season for that one experimental book that editorial can agree is so crazy it just might work? The rest of the list can be meat and potatoes, but once a season a book should make it to publication that will either blaze a trail to the stars or end up as a beautiful pillar of smoke at the end of the runway. Otherwise, why bother?
Likewise, readers should get their sense of adventure back. Read something freaky once a year, why don't you? You could read something that you don't connect with at all, or you could wind up having a revelation. Either way, at least you won't be bored, and you'll be saving the publishing industry from itself.